logoalt Hacker News

glimshe • today at 11:46 AM • 5 replies • view on HN

Why didn't Claude detect the hack and prevent it from affecting customers using its superhuman AGI reasoning?


Replies

_heimdall • today at 11:50 AM

Don't worry, I assume an LLM version of McAfee Total Protection Antivirus is coming soon

➕ show 2 replies
viraptor • today at 1:34 PM

This is a cheap troll, but just in case someone is wondering: This extension likely never hit the standard workflow and/or repository. Most likely someone got the credentials and uploaded a locally modified version. No review system at anthropic has seen this version.

➕ show 4 replies
MisterTea • today at 12:35 PM

They weren't using last minutes model, which is totally better than last hours model which would have totally detected this, pinkie swear.

➕ show 2 replies
eviks • today at 11:50 AM

It was safety-contrained from being able to realize its full potential.

Hackbraten • today at 12:47 PM

You’re absolutely correct, and I owe you a honest statement—it was a hack, and here’s the kicker: *I failed to catch it, and that’s on me.*

Here’s a complete rundown of what happened—load-bearing parts first: